Publicación: La influencia de la Inteligencia Artificial Generativa en la Ciberseguridad: Deepfakes
Cargando...
Fecha
2024-06
Autores
Editor/a
Director/a
Tutor/a
Coordinador/a
Prologuista
Revisor/a
Ilustrador/a
Derechos de acceso
info:eu-repo/semantics/openAccess
Título de la revista
ISSN de la revista
Título del volumen
Editor
Resumen
La Inteligencia Arti cial ha avanzado mucho desde su inicio en 1943, cuando se propuso el primer modelo que simulaba la actividad de una red neuronal. Desde entonces, se han llevado a cabo muchas investigaciones que han llevado a la tecnología al punto de partida del que hoy en día se encuenta y desde el que parece que puede llegar a cambiar e in uir en nuestra vida de un modo indiscutible. Cabe destacar el trabajo de Goodfellow en 2014, cuando publicó su importante trabajo sobre la Inteligencia Arti cial desde un punto de vista totalmente distinto.
En la década de 1960, la Inteligencia Arti cial Generativa había comenzado a emerger con chatbots. Estos no estaban demasiado desarrollados, aunque cumplían con su función. Good-fellow publicó entonces un modelo que era capaz de poder generar imágenes, vídeos y audios sintéticos. A este modelo se le bautizó como Redes Generativas Adversarias.
Las aplicaciones de este modelo demostraron posteriormente ser muy numerosas: automatización de la generación de arte para redes sociales, pancartas o infografías; canciones generadas sintéticamente o incluso vídeos vídeos sin tener si quiera que ser grabados. Sin embargo, entre las aplicaciones también se puede encontrar alguna con nes maliciosos. Esta tecnología puede permitir suplantar identidades vulnerando principios tales como la integridad o la privacidad y suponer un riesgo muy alto en la ciberseguridad.
Este Trabajo de Fin de Máster tiene como objetivo enumerar estos riesgos, centrándose principalmente en los deepfakes y la tecnología que los soporta. También se enumeran otras aplicaciones en ciberseguridad.
Arti cial Intelligence has come a long way since its inception in 1943, when the first model simulating the activity of a neural network was proposed. Since then, much research has been carried out that has brought technology to the starting point where it is today and from where it seems that it can change and in uence our lives in an indisputable way. Noteworthy is the work of Goodfellow in 2014, when he published his important work on Arti cial Intelligence from an entirely di erent point of view. In the 1960s, Generative Arti cial Intelligence had begun to emerge with chatbots. These were not very developed, although they ful lled their function. Goodfellow then published a model that was capable of generating synthetic images, videos and audio. This model was named after the name of Adversary Generative Networks. The applications of this model later proved to be very numerous: automation of the generation of art for social networks, banners or infographics; synthetically generated songs or even videos without having to be recorded. However, among the applications you can also and some with malicious purposes. This technology can make it possible to impersonate identities in violation of principles such as integrity or privacy and pose a very high risk to cybersecurity. This Thesis aims to list these risks, focusing primarily on deepfakes and the technology that supports them. Other cybersecurity applications are also listed.
Arti cial Intelligence has come a long way since its inception in 1943, when the first model simulating the activity of a neural network was proposed. Since then, much research has been carried out that has brought technology to the starting point where it is today and from where it seems that it can change and in uence our lives in an indisputable way. Noteworthy is the work of Goodfellow in 2014, when he published his important work on Arti cial Intelligence from an entirely di erent point of view. In the 1960s, Generative Arti cial Intelligence had begun to emerge with chatbots. These were not very developed, although they ful lled their function. Goodfellow then published a model that was capable of generating synthetic images, videos and audio. This model was named after the name of Adversary Generative Networks. The applications of this model later proved to be very numerous: automation of the generation of art for social networks, banners or infographics; synthetically generated songs or even videos without having to be recorded. However, among the applications you can also and some with malicious purposes. This technology can make it possible to impersonate identities in violation of principles such as integrity or privacy and pose a very high risk to cybersecurity. This Thesis aims to list these risks, focusing primarily on deepfakes and the technology that supports them. Other cybersecurity applications are also listed.
Descripción
Categorías UNESCO
Palabras clave
Inteligencia artificial, GenAI, Ciberseguridad, Deepfakes, Artificial Inteligence, Cybersecurity
Citación
López Delgado, Juan Luis. "La influencia de la Inteligencia Artificial Generativa en la Ciberseguridad: Deepfakes". Dirigido por Pastor Vargas, Rafael, Trabajo Fin de Máster. Universidad Nacional de Educación a Distancia, Escuela Técnica Superior de Ingeniería Informática (2024)
Centro
E.T.S. de Ingeniería Informática
Departamento
Sistemas de Comunicación y Control